Effects of West Nile Virus on behavioral and cognitive performance, circadian body temperatures, cortical Aβ and p-tau pathology, and immune measures in human apoE and human amyloid precursor protein NL-G-F/human apoE mice.

Introduction: There are three human apolipoprotein E (E) isoforms: E2, E3, and E4. Compared with E3, E4 is an AD risk factor. West Nile Virus (WNV) can cause neurological disease, and it results in some neuropathology and neuroinflammation seen in Alzheimer's disease (AD). Previously, we reported on the effects of exposure to the Kunjin strain (KUNV) of WNV in middle-aged knock-in (KI) mice expressing human amyloid precursor protein (APP) containing the dominant NL-G-F mutations with human apoE targeted replacement (TR) mice on behavioral, immune, and pathological measures, and circadian body temperatures. NL-G-F/E4 mice were more susceptible to KUNV-induced cognitive injury and prolonged viral load in the cortex; viral load was detected in the cortex of more mice expressing amyloid precursor protein with the dominant NL-G-F AD mutations and expressing also E4 than those expressing also E3. As NL-G-F and NL-G-F/apoE mice show cognitive injury and neuropathology at 6 months of age and age is a risk factor to develop AD and modulates the immune response in AD, we assessed in this study whether genotype differences in response to KUNV exposure are seen in younger mice. To determine whether the presence of NL-G-F is required to detect genotype differences in response to KUNV and/or modulates the response to KUNV, E3 and E4 mice were included as well. As compared with E3, E2 reduces the risk to develop AD and the vast majority of AD would likely not occur in the absence of E3 and E4 NL-G-F/E2 mice were included in this study as well. Methods: We crossed knock-in (KI) mice expressing human amyloid precursor protein (APP) containing the dominant NL-G-F mutations with human apoE targeted replacement (TR) mice and used middle-aged NL-G-F/E3 and NL-G-F/E4 mice to assess the role of WNV (subtype Kunjin virus) (KUNV) exposure on behavioral alteration performance, circadian body temperatures, viral loads, neuropathology, and transcript levels of three immune measures in cortex important in the detrimental effects of KUNV on brain function. Results: There were apoE isoform-dependent responses to KUNV exposure on various measures already seen at 6 months of age and are mostly more profound in the absence than in the presence of NL-G-F. One week following exposure, cortical insoluble p-tau levels were lower in KUNV- than PBS-treated apoE, but not NL-G-F/apoE mice. Soluble p-tau levels in cortex were higher in KUNV- than PBS-treated mice, and this effect was more pronounced in NL-G-F/apoE than apoE mice. The insoluble/soluble p-tau ratio was lower in KUNV- than PBS-treated mice, and this effect was more pronounced in apoE than NL-G-F/apoE. Discussion: These results are consistent with results in cultured cells supporting that Aβ and p-tau might have antiviral properties.